The presence of gravel in soils modifies the porosity, pore connectivity and pore size distribution in the soil matrix as well as the soil matrix-gravel interfaces. The aim of the present study is to investigate the effect of relative volume of gravel in samples with gravel mass fractions of 5,10, 20 wt% and varying bulk densities (1.3, 1.45, 1.55, 1.60, 1.65 g cm–3) on (i) total porosity, field capacity, plant available water holding capacity, (ii) pore size distribution and (iii) thermal capacity of repacked sandy and silty soils. The focus of the study was to determine if laboratory measured soil water retention curves considering (i), (ii), and (iii) can be predicted by a gravel-based weighting factor, Rv, considering comprehensive significance tests. The sand-gravel mixtures show a decrease in the volume fractions of macropores and wide cores pores with an increase in the gravel contents, while the silt-gravel mixtures show an opposite trend. The root mean square errors (RMSE) between measured and fitted volumetric water contents, θ, between 0.006 and 0.0352 and between 0.002 and 0.004 for Rv-weighted volumetric water contents indicate that the van Genuchten-based Peters-Durner- Iden (PDI) model is appropriate for fitting. The soil water retention curves with mass gravel contents of up to 10 wt% for silt and 20 wt% for sand can be well predicted by weighting factors (relative volume of rock fragments) in the range between 0.045 and 0.058 for silt, and between 0.112 and 0.119 for sand. The results also indicate a decrease in the Rv-weighted saturated, cvsat, and dry, cvdry, thermal capacity with an increase in the gravel contents for both soils. Further investigations are needed to examine if and whether measured sand- and silt-gravel mixtures with mass gravel contents below 10 % or rather 20 % can be predicted with a weighting factor.
The idea of the present study is to describe the spatially varying particle size distribution (PSD) along intact aggregate surfaces with the laser diffraction method (LDM) of four silty-loamy and OC enriched horizons of a Dystric Cambisol from the Uhlířská catchment (Czech Republic) with the laser diffraction method (LDM). Besides, the comparability of the LDM with the sieve and pipette method (SPM), the reproducibility, and the effect of pretreatment on the particle size distribution derived by LDM were analysed. The laser diffraction method enables rapid and continuous particle size distribution measurements with required sample amounts of 0.1–0.2 g for each measurement compared to 5–20 g for SPM. The LDM-derived PSD’s can be directly compared with the standardised SPM-derived PSD’s by using regression analysis with coefficients of determination (r²) between 0.83 and 0.93. Sample pretreatment following standardised proceedings indicates a better comparability between the particle size distributions of both methods. Besides, the highest coefficients of variation of up to 78.6 and therefore the lowest reproducibility were found for the unpretreated PSD of the AE and Bs horizon. Thus, limited evaluability and reproducibility of soil material enriched in organic carbon (OC), used in the current study, needs further analysis. For spatial analysis of PSD’s along intact surfaces of soil aggregates and profiles, spatial data interpolation by inverse distance weighting (IDW), kriging, and triangulated irregular networks (TIN) can be used for detailed measuring, mapping, and spatial extension of the sand, silt, and clay fractions at unsampled locations using a set of samples of known locations. The information offers the possibility of comparing and verifying data obtained by non-invasive midinfrared spectroscopy and Vis–NIR spectroscopy by spatial extension for given soil aggregates and profiles.